The Agent Skills Directory

[SAFE]: The skill facilitates the generation of visualization code using established Python libraries including matplotlib, seaborn, and numpy. These are standard tools for data science and are retrieved from official registries.
[SAFE]: It implements a secure validation mechanism using the Python AST (Abstract Syntax Tree) module. The command python -c "import ast; ast.parse(...)" allows the agent to verify the syntax of the generated code without executing the script itself, preventing accidental runtime execution of untrusted content during the check.
[SAFE]: No patterns of data exfiltration, credential harvesting, or malicious obfuscation were detected in the instructions or metadata.
[SAFE]: The instructions for the user to execute the generated code are transparent and consistent with the skill's stated purpose of producing runnable research figures.

figure